Grid Integration of Wind Power

advertisement
Grid Integration of Wind Power
Best Practices for Emerging Wind Markets
Issues with grid integration of wind energy has led to curtailment of wind power, delay in interconnection for
commissioned wind projects and/or denial of generation permit. This report describes the impact of wind power
on the grid, methods to analyze the impact and approaches to mitigate the impact. Countries like Denmark,
Germany, Spain, and regions within the United States like Texas and Colorado have achieved high penetration
of wind energy with modest changes to the grid. The key to high penetration of wind energy has been flexible
grids. The report outlines the lessons learned with a focus on applicability to emerging wind energy markets.
GRID INTEGRATION
OF WIND POWER
About the Asian Development Bank
ADB’s vision is an Asia and Pacific region free of poverty. Its mission is to help its developing member countries
reduce poverty and improve the quality of life of their people. Despite the region’s many successes, it remains
home to the majority of the world’s poor. ADB is committed to reducing poverty through inclusive economic
growth, environmentally sustainable growth, and regional integration.
BEST PRACTICES FOR EMERGING
WIND MARKETS
Based in Manila, ADB is owned by 67 members, including 48 from the region. Its main instruments for helping
its developing member countries are policy dialogue, loans, equity investments, guarantees, grants, and
technical assistance.
Pramod Jain and Priyantha Wijayatunga
NO. 43
April 2016
ASIAN DEVELOPMENT BANK
6 ADB Avenue, Mandaluyong City
1550 Metro Manila, Philippines
www.adb.org
ADB SUSTAINABLE DEVELOPMENT
WORKING PAPER SERIES
ASIAN DEVELOPMENT BANK
i
Sustainable Development Working Paper Series
Grid Integration of Wind Power: Best Practices
for Emerging Wind Markets
Pramod Jain and Priyantha Wijayatunga
No. 43 | April 2016
Priyantha Wijayatunga is principal energy
specialist, Sustainable Development and
Climate Change Department, Asian
Development Bank (ADB).
Pramod Jain is the president of Innovative
Wind Energy and consults with ADB
under the Quantum Leap in Wind
Development for Asia and the Pacific
Technical Assistance program.
Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO)
© 2016 Asian Development Bank
6 ADB Avenue, Mandaluyong City, 1550 Metro Manila, Philippines
Tel +63 2 632 4444; Fax +63 2 636 2444
www.adb.org; openaccess.adb.org
Some rights reserved. Published in 2016.
Printed in the Philippines.
Publication Stock No. WPS167996-2
Cataloging-In-Publication Data
Asian Development Bank.
ADB Sustainable Development Working Paper Series No. 43: Grid Integration of Wind Power: Best Practices for
Emerging Wind Markets.
Mandaluyong City, Philippines: Asian Development Bank, 2016.
1. Wind energy.
2. Renewable energy.
I. Asian Development Bank.
The views expressed in this publication are those of the authors and do not necessarily reflect the views and policies of the
Asian Development Bank (ADB) or its Board of Governors or the governments they represent.
ADB does not guarantee the accuracy of the data included in this publication and accepts no responsibility for any
consequence of their use. The mention of specific companies or products of manufacturers does not imply that they are
endorsed or recommended by ADB in preference to others of a similar nature that are not mentioned.
By making any designation of or reference to a particular territory or geographic area, or by using the term “country” in this
document, ADB does not intend to make any judgments as to the legal or other status of any territory or area.
This work is available under the Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO)
https://creativecommons.org/licenses/by/3.0/igo/. By using the content of this publication, you agree to be bound by the
terms of said license as well as the Terms of Use of the ADB Open Access Repository at openaccess.adb.org/termsofuse
This CC license does not apply to non-ADB copyright materials in this publication. If the material is attributed to another
source, please contact the copyright owner or publisher of that source for permission to reproduce it. ADB cannot be held
liable for any claims that arise as a result of your use of the material.
Attribution—In acknowledging ADB as the source, please be sure to include all of the following information:
Author. Year of publication. Title of the material. © Asian Development Bank [and/or Publisher].
https://openaccess.adb.org. Available under a CC BY 3.0 IGO license.
Translations—Any translations you create should carry the following disclaimer:
Originally published by the Asian Development Bank in English under the title [title] © [Year of publication] Asian
Development Bank. All rights reserved. The quality of this translation and its coherence with the original text is the sole
responsibility of the [translator]. The English original of this work is the only official version.
Adaptations—Any adaptations you create should carry the following disclaimer:
This is an adaptation of an original Work © Asian Development Bank [Year]. The views expressed here are those of the
authors and do not necessarily reflect the views and policies of ADB or its Board of Governors or the governments they
represent. ADB does not endorse this work or guarantee the accuracy of the data included in this publication and accepts no
responsibility for any consequence of their use.
Please contact [email protected] or [email protected] if you have questions or comments with respect to content, or
if you wish to obtain copyright permission for your intended use that does not fall within these terms, or for permission to use
the ADB logo.
Note: In this publication, “$” refers to US dollars.
CONTENTS
TABLES AND FIGURES
iv
ABBREVIATIONS
v
EXECUTIVE SUMMARY
vi
I.
INTRODUCTION
1
II.
WHAT IS UNIQUE ABOUT INTEGRATING WIND ENERGY?
4
A.
Variable Power
4
B.
Uncertain Power
4
C.
Geographic Diversity, Size, and Distance from Load
4
D.
Standardized Power Purchase Agreement with Guaranteed Interconnection
and Priority Dispatch
5
E.
Wind Turbine Generators
6
III.
IV.
V.
WHAT IS GRID INTEGRATION OF WIND ENERGY?
6
A.
Planning
7
B.
Physical Connection
8
C.
System Operations
9
WHAT IS THE IMPACT OF A WIND POWER PLANT ON A GRID?
9
A.
Day-Ahead Unit Commitment Process
9
B.
Economic Dispatch Process
9
C.
Illustration of Variability and Uncertainty of Wind Energy
10
D.
Implications of Wind Energy on Conventional Generators and Transmission
10
BEST PRACTICES FOR WIND INTEGRATION IN EMERGING MARKETS
11
A.
Transmission from Resource-Rich Areas to Load Center
12
B.
Responsive System Operations
12
VI.
HOW MUCH DOES GRID INTEGRATION COST?
15
VII.
CONCLUSIONS
18
REFERENCES
20
TABLES AND FIGURES
Tables
1 Common Myths about Grid Integration of Wind Energy
2
Figures
1 Plot of Net Load for 1 Week
3
2 High-Level Activities of Grid Integration
6
3 Tiered Planning for Grid Integration
15
4 Studies on Wind Integration Costs versus Wind Capacity Penetration in Various Regions
of the United States
16
5 Relative Cost of Various Options to Increase Flexibility of the Grid with Varying
Wind Percentages
17
ABBREVIATIONS
ADB
FiT
GW
MW
MWh
POI
VRE
WPP
WTG
–
–
–
–
–
–
–
–
–
Asian Development Bank
feed-in tariff
gigawatts
megawatts
megawatt-hour
point of interconnection
variable renewable energy
wind power plants
wind turbine generators
vi
Executive Summary
EXECUTIVE SUMMARY
Utility-scale wind energy is the most inexpensive form of renewable energy in countries with good wind
resources. Globally, wind power has experienced rapid growth of 25% annually, from 17.4 gigawatts
(GW) in 2000 to 370 GW in 2014. Growth in Asia has been limited to the People’s Republic of China
and India. As of end of 2014, the former had 114.8 GW, which is 31% of the global installed wind
capacity, while India had 22.5 GW, for a 6.1% share. Other Asian countries like Pakistan, the Philippines,
and Thailand have seen a growth spurt in 2013 and 2014.
The Quantum Leap in Wind Power Development for Asia and the Pacific program of the Asian
Development Bank (ADB) has been providing technical assistance for close to 5 years to three
emerging wind energy countries—Mongolia, the Philippines, and Sri Lanka. This study reports
observations and experiences in these and other emerging wind energy markets.
In emerging wind energy markets, issues related to grid integration of wind energy are not at the
forefront. Grid integration is approached on an ad hoc, project-by-project basis. Project-specific grid
integration components like transmission, substation, and physical connection are built in order to
transport wind energy from a wind power plant (WPP) to load centers. Beyond this obvious
component, all other grid integration components are ignored until after the WPP is online and issues
start to emerge like power quality (harmonics), voltage fluctuations, congestion, and curtailment.
There is also a pervasive notion among grid operators that their specific grid is unique and there is not
much that can be learned from other grid operators, especially from developed wind energy markets.
Features of Wind Energy
Wind energy (and other forms of variable renewable energy like solar photovoltaic cells) has the
following unique characteristics, and consequently, implications when connected to the grid:
(i)
Wind energy is variable and uncertain. The first implication is that other generators on the
grid must react not only to load variability, but also to wind energy variability. This requires
conventional generators to support higher ramp rates in both directions. The second
implication is that for cost-efficient operations, accurate wind energy forecasts and
responsive systems operations are necessary.
(ii) Wind generator is asynchronous in contrast to synchronous conventional generators. The
implication is that wind energy provides no inertial response, so when wind energy
penetration is high, grid inertia can be low. As a consequence, the grid may become
unstable—unable to provide a stable subsecond-to-second frequency response in
reaction to fault on the grid.
(iii) Wind energy is on priority dispatch. The first implication is that it displaces conventional
generation, leading to lower minimum operating levels, and longer and more frequent
shutdowns of conventional generators. The second implication is that wind energy may be
curtailed during periods of off-peak load, high wind production, and generators operating
at minimum capacity.
Executive Summary
vii
(iv) Wind energy is usually not proximate to load centers. The implications are new transmission
or upgrade to existing transmission may be required; losses may be high; and voltage level
may fluctuate beyond acceptable levels at interconnection point, thereby necessitating
reactive power compensation.
Best Practices for Emerging Wind Energy Markets
Against this backdrop, the ad hoc project specific approach to grid integration is grossly inadequate to
support sustainable wind power development. The following best practices are derived from grid
integration failures and successes in the pioneering and mature markets.
(i)
Transmission from resource-rich areas to load centers. Since transmission projects have a
longer lead time than wind projects, transmission projects, whether new or upgrade, must
be planned ahead of time. A policy of defining wind energy corridors and focusing
transmission investment in a corridor is preferable to a project-specific approach.
(ii) Responsive systems operations. Upgrading the scheduling and dispatching process for all
generators to sub-hour dispatching intervals, incorporating sub-hour lead time for wind
energy forecasts, and expanding the balancing areas are process-related improvements
that offer the cheapest form of grid flexibility.
(iii) Flexible generation. Variability in a grid is not new; it is managed by flexible generation.
Wind energy introduces additional variability that must be served by generators
with higher ramp rates and deeper cycles. If low grid inertia is an issue, then to support
high wind penetration, wind turbines should be required to provide inertia and
governor-like response.
(iv) Flexible demand. Wind curtailments occur during periods when wind production is higher
than the difference between the load and sum of minimum operating levels of all
dispatched generators. Load shifting to such periods from periods of peak load can
enhance the ability of the grid to absorb a higher amount of variable energy.
(v) Comprehensive planning process. There is no one-size-fits-all solution that applies to all
grids. A network-wide grid integration impact assessment that examines the impact on
the grid of various percentages of wind energy penetration as a function of time is
required to determine the grid integration issues and solutions.
(vi) Grid code for wind integration. A grid code for connecting wind plants is essential because it
specifies the minimum technical criteria that all wind farms seeking connection to the grid
shall satisfy at the point of interconnection (POI), and the operating conditions that all
wind farms must comply with for safe and reliable operations of the grid.
viii
Executive Summary
Country Experiences with Integrated Wind Energy
Experiences from other countries are exemplified in the following observations in a 2015 Wind Vision
report (US Department of Energy 2015, chapter 2.7):
(i)
The electric power network operated reliably with high wind contributions (10% or higher)
in 2013 with minimal impacts on network operating costs.
(ii) In regions with wind power contributions up to 20% of annual electrical demand in 2013,
electric power systems operated reliably without added storage and with little or no
increase in generation reserves.
(iii) Wind has been proven to increase system reliability during some severe weather events.
(iv) Studies prior to 2008 had estimated integration costs up to $5 per megawatt-hour
(MWh). By 2013, the Electric Reliability Council of Texas, which has the highest
penetration of wind, was reporting integration cost of $0.5/MWh.
Although costs of wind integration are grid-specific, large numbers of studies have indicated that the
cost of integration for low levels of wind energy penetration (less than 5%) is negligible, and a higher
level of wind penetration (20% or higher) is less than 10% of the cost of wind energy. The question is
no longer “can wind energy be integrated?” but “how should wind energy be integrated?”
I.
INTRODUCTION
Wind energy is widely recognized as one of the cheapest forms of clean and renewable energy. In fact,
in several countries, wind energy has achieved cost parity with fossil-fuel-based sources of electricity
generation for new electricity generation plants (McCrone et al. 2014). The focus of this publication is
on grid integration issues in emerging wind energy markets. Although the content of this publication is
applicable to all variable renewable energy sources like wind, solar, tidal and hydro, the focus will be on
wind energy. 1
The input raw material (feedstock) for a wind power plant (WPP) is wind; therefore, there is no
pollution or environmental degradation due to mining and/or transporting of raw materials like coal,
crude oil, or natural gas. The output is clean electricity with absolutely no pollutants—no carbon
dioxide, sulfur oxides, or nitrogen oxides—and the WPPs do not use water. All these favorable
characteristics of WPPs provide significant tangible benefits in terms of improved health and
conservation of natural resources.
Wind energy is a proverbial three-legged stool. There is no wind project without these three legs:
(i)
Wind resources. Without good wind resources, there is no fuel, hence no project.
(ii) Power purchase agreement with a tariff. Without a reasonable tariff and a signed contract to
purchase all, a wind project is not economically feasible and there will be no investment.
(iii) Grid integration. Without interconnection of a WPP to the grid, there is no method to
deliver electricity to the buyers.
Through wind resource assessment, private wind developers choose locations with good wind
resources. In most emerging wind energy markets, the feed-in tariff (FiT) policy is prevalent, which
guarantees a standard power purchase agreement, published tariff, guaranteed interconnection to the
grid, and priority dispatch. A well-designed FiT policy therefore takes care of the other two legs.
Just because an interconnection is guaranteed, it does not mean the interconnection process will be
smooth, prompt, or inexpensive. In most emerging wind markets, grid integration is an afterthought.
This has sometimes led to delayed interconnection (wind plants stay idle for months after
commissioning) and/or significant amount of wind energy curtailment. Table 1 lists some common
myths of wind energy related to grid integration. Figure 1 illustrates the impact of wind on variability of
net load.
1
Several individuals assisted in seeding the idea, approving the work on the report, reviewing the report, and above all
providing constructive criticism. Bo An, public management specialist in the East Asia Department of the Asian
Development Bank (ADB) proposed the idea of undertaking a study on grid integration of wind power with a focus on
developing wind power markets. Aiming Zhou, energy specialist in the South Asia Department (SARD) reviewed an early
draft of the report and provided guidance on its structure. Priyantha Wijayatunga, principal energy specialist, Sustainable
Development and Climate Change Department, ADB; and Kazuhiro Enomoto, energy specialist, SARD, provided valuable
insights and led the effort of preparing the study for publication.
2
ADB Sustainable Development Working Paper Series No. 43
Table 1: Common Myths about Grid Integration of Wind Energy
Wind Energy Myths Related
to Grid Integration
Reality
Variability and uncertainty of
wind cannot be managed by a
grid; it is too disruptive
Electricity grids are designed to manage both variability and uncertainty of loads,
including occasional failures of one or more generation units, transmission, and
substation (N-1 scenario).
To explain why variable renewable energy (VRE) generation, which includes wind and
solar, is not much more disruptive than normal load variation, consider the following.
Often for modeling purposes VRE generation is considered as negative load. That is, VRE
generation is subtracted from load to compute net load. The net load may have slightly
higher variability and slightly higher uncertainty compared to load from consumer
demand. This should be analyzed specifically for the grid under analysis.
As an example, consider the case of Texas with 15,000 megawatts (MW) of wind. The
variability added by wind energy for different time periods was negligible: (i) 6.5MW
(0.04%) for 1 minute; (ii) 30MW (0.2%) for 5 minutes; and (iii) 328MW (2.2%) for
1 hour, where 1 minute, 5 minutes, and 1 hour are averaging periods (American Wind
Energy Association). As a second example, consider the case of Western Denmark,
where the maximum hourly wind power swing was 18% of installed capacity, and for 50%
of the time it was below 2% of installed capacity (International Energy Agency. 2005).
Managing this additional variability and uncertainty is not radically different compared to
capabilities of current grids. More than 30 years of experience in the People’s Republic of
China, the European Union, India, the United States, and other areas has shown that
VRE can be managed and can be done at modest cost.
Wind generation can drop to
zero in seconds
Small disturbances in grid can
cause wind plants to trip,
causing cascading failure and
total collapse of the grid
WPPs do not cause grid collapse and cascading failure. Wind speed does not suddenly
drop and multiple wind turbines all do not see the same wind speed. When the
generation output of multiple turbines is aggregated, it is observed that wind generation
output falls smoothly. This is unlike solar plants where the output can fall quickly as
clouds arrive.
Most grid codes for interconnection of WPP require low-voltage ride through
capability, which prevents the wind farm from disconnecting due to a transient fault
(short disturbance). In fact studies have shown that modern wind power plants are
grid-friendly (Ela et al. 2014). For example, GE wind turbines provide no trip during
transient faults or disturbances, regulate plant voltage and reactive power, react to
changes in grid frequency, provide inertial response to large under-frequency events,
and limit the ramp rate (Millet et al. 2013).
Wind energy causes excessive
cycling of thermal plants,
which significantly increases
the cost of thermal generation
Studies have shown that the increase in cost of thermal generation is minimal. NREL’s
Western Wind and Solar Integration Study found that with 33% wind and solar energy
(25% wind, 8% solar), the additional cost of operation and maintenance of thermal
plants is $0.47 to $1.28 per megawatt-hour (MWh), (Jordan and Venkataraman 2012)
while the overall reduction in system operating costs of thermal plants (fuel savings and
others) is $85/MWh.
Although WPP do cause additional cycling of conventional generators, the additional
cost is minimal.
Wind energy causes excessive
cycling of thermal plants,
which increases GHG
emissions
Studies have shown that the GHG emissions increase due to cycling of thermal plants is
minimal. With 33% VRE (National Renewable Energy Laboratory 2013), carbon dioxide
emissions are reduced by 29%–34% due to VRE, while cycling has a negligible impact on
this reduction. Sulfur dioxide emissions are reduced by 14%–24% due to VRE with
cycling, reducing the benefit by 2%–5%. Nitrous oxide emissions are reduced by
16%–22% due to VRE, with cycling increasing the benefit by 1%–2%.
GHG emissions reductions due to WPPs are marginally impacted by cycling of
conventional power plants.
continued on next page
Grid Integration of Wind Power: Best Practices for Emerging Wind Markets
3
Table 1 continued
Wind Energy Myths Related to
Grid Integration
Wind energy imposes high cost
in terms of reserves and storage
a
Reality
Reserves are always managed at the system level and not at the level of individual
generation or type of generation. So 1 MW of reserve for 1 MW of wind is a myth.
Furthermore, studies have shown that for most grids in the United States, storage is
not needed for up to 30% of wind penetration (Cochran et al. 2014); even beyond
that, storage is one of many options, and others are lower-cost options. However, for
a
inflexible grids, reserves and storage may be required at lower levels of penetration.
Large integration grid studies have indicated that even with 33% penetration of wind
energy, the increase in cost is relatively low ($5/MWh). Small isolated systems have
higher costs, while for large, diverse, and agile systems there may be little increase in
cost. According to the American Wind Energy Association, on average adding 3 MW
of wind energy to the US grid would result in a modest 0–0.01 MW of additional
spinning reserves, and 0–0.07 MW of nonspinning reserves.
Flexible grids are described in Section V. For more details see Cochran et al. 2014.
Note: The research presented here is specific to the studied grid. Although the results should translate to other grids, a
grid-specific study should be performed.
Source: Asian Development Bank.
Figure 1: Plot of Net Load for 1 Week
Net load = load – wind.
Note: Net load illustrates the impact of wind on variability. Variability is measured in terms of megawatts per unit time,
where time may be one minute, 10 minutes, or one hour. Steeper ramps and lower turn-down are examples of higher
variability of net load. Since the conventional generation responds to net load, it sees higher variability, which is reflected in
requirements for higher ramp and turn-down rates.
Source: Cochran et al. 2014.
4
ADB Sustainable Development Working Paper Series No. 43
The sections of this report are organized to answer specific questions. The second section answers the
question, What is unique about wind energy with respect to grid integration? The third section answers
the question, What are the components of grid integration of a WPP? The fourth section answers what
is the impact of wind energy on a grid. The fifth section addresses what are the best practices for grid
integration of wind energy. The sixth section answers the question what is the cost of wind integration.
Section VII concludes.
II.
WHAT IS UNIQUE ABOUT INTEGRATING WIND ENERGY?
This section describes the unique characteristics of wind resource, wind energy policies, and wind
turbine generators.
A.
Variable Power
The energy output of WPPs is variable because the wind speed is variable. Variability is in all time
scales—hour-to-hour, day-to-day, month-to-month, and year-to-year variability. Variability and
uncertainty are very different concepts and are often misunderstood. Variability in wind speed is
caused by the earth’s rotation (day–night cycle), tilt axis of the earth (seasons), and other natural
phenomenon. Examples of variability are higher wind speeds occurring in early morning hours and late
evening, and higher wind speeds happening in spring and autumn. Hour-to-hour, day-to-day, and
month-to-month changes in wind speed cause variability in wind energy production. A conventional
base-load thermal power plant may be considered to have minimal variability as long as sufficient fuel
is delivered to it; any variability is due to planned activities like scheduled maintenance. Among variable
renewable energy sources, solar dominates intraday variability because of sunrise and sunset. When
diurnal variability is removed, then solar and wind have similar variability (Lew et al. 2013).
B.
Uncertain Power
Uncertainty has to do with unpredictability of wind speed. Continuing with the hypothetical case, the
randomness around the variable pattern of wind speed is the uncertainty. Since wind is a weather
phenomenon, uncertainty occurs in all time-scales, year-to-year, month-to-month, day-to-day,
hour-to-hour, and smaller. A variety of forecasting methods are used to predict wind energy
production. The following are true about accuracy of wind energy forecasts (Jain 2010): (i) day-ahead
forecast is less accurate than hour-ahead forecast, which is less accurate than 15-minute-ahead
forecast; (ii) forecast accuracy increases as the number of turbines increases in a WPP; and
(iii) forecast accuracy increases as the number of WPPs increase and the WPPs are in geographically
diverse locations. Among renewable energy sources, wind energy has higher uncertainty compared
to solar.
C.
Geographic Diversity, Size, and Distance from Load
Wind power plants are located where the wind resource is high. In most parts of the world, areas with
high wind resources are far away from population centers. The reason is it is difficult to live in areas
with sustained high wind speeds. Such areas are either not served by a central grid or served by a
“weak” grid. A weak grid is characterized by a long transmission line that carries a small amount of
power. An implication for wind projects in such areas is the need to build new or upgrade existing
transmission lines. Long transmission lines from WPP to load center may lead to fluctuation in voltage
Grid Integration of Wind Power: Best Practices for Emerging Wind Markets
5
that is larger than the allowable limit specified in the grid code. This should be managed by using
reactive power compensators on the line.
The second geographical aspect of WPPs is its density. The density of a wind farm is typically 5 MW to
8 MW per square kilometer. For example, a 50 MW WPP would be spread across 6–10 square
kilometers. The output of a WPP is an aggregate over multiple turbines, and if there are multiple WPPs
in the grid, then wind energy injected into the grid is an aggregate over the WPPs. Aggregation
contributes to reduction in uncertainty of wind energy (because all turbines do not see the same wind
speed profile) and to a lesser extent reduction in variability of wind energy.
The other implication of modularity of WPP is that scheduled or unscheduled maintenance of one
turbine does not materially impact the output of the WPP. Hence, in the aggregate, a WPP can exhibit
higher reliability.
D.
Standardized Power Purchase Agreement with Guaranteed Interconnection
and Priority Dispatch
In most emerging wind energy markets, wind energy projects avail of the following incentives:
(i)
Standardized power purchase agreement (SPPA). Most emerging wind energy countries
have a single-buyer market. SPPA is intended to simplify or in most cases eliminate the
need for negotiating with the off-taker (buyer of wind energy).
(ii) Feed-in tariff. FiT is a cost-based tariff that is intended to ensure a reasonable return on
investment for a wind power developer.
(iii) Priority dispatch. FiT is often accompanied with take-or-pay contract in which the wind
farm owner is paid even for curtailed energy. Curtailments occur when the wind energy
production is higher than the difference between the load and sum of minimum
production levels of all dispatched generators. Details of take-or-pay contracts are
country-specific, but except in cases of force majeure, safety of grid, or reliability of
supply, the WPP owners are paid for amount of curtailed energy. Therefore, it is in the
economic interest of the grid to absorb all energy produced by WPPs. In addition, since
wind energy has near-zero marginal cost (because there is no fuel cost), it should be
considered as a priority dispatch resource. In several competitive electricity markets
this sometimes manifests as negative prices during times of high wind production and
low demand.
(iv) Guaranteed interconnection. SPPA includes provisions for guaranteed interconnection of
WPPs to the grid as long as the WPP meets provisions in the grid code. In most countries,
the cost of building transmission from the WPP to an existing transmission line is borne by
the WPP, while the costs of upgrading the substation and transmission line are subject to
cost sharing with the transmission company. If transmission upgrade is planned to be
exclusively used by the proposed WPP, then the WPP pays for the upgrade. If it is
recognized that the upgrades have a regional benefit and it can be included in a regional
transmission plan, then cost of upgrades may be shared.
6
ADB Sustainable Development Working Paper Series No. 43
(v) Demand-side incentives. FiT is a supply-side incentive provided to the seller of wind energy.
In cases where FiT is higher than the prevailing tariff paid to the generators displaced by
wind power (marginal tariff), then the buyer is compensated for the difference between
FiT and the marginal tariff. Typically, renewable energy funds (REF) are created to
compensate the buyer.
These incentives give the incumbent generation companies and the transmission company a mistaken
perception that wind energy generation is getting preferential treatment. What is forgotten is that all
technologies in a nascent phase get preferential treatment, and as a society, clean technologies should
get preferential treatment.
E.
Wind Turbine Generators
Wind turbine generators (WTG) technology has evolved from Type 1 and 2 induction-based
generators to Type 3 double-fed induction generator and Type 4 full-power conversion-based
generator. Type 1 and 2 generators consumed reactive power from the grid, therefore were not
grid-friendly and by and large are no longer used in utility scale turbines. Type 3 and 4 WTGs are
grid-friendly and can consume or produce reactive power to support grid functions. These turbines
provide active and reactive power control, low-voltage ride through, high-pass filters, and other
features that can increase the stability and reliability of the grid.
Wind turbine generators do not provide inertial and governor response, like conventional synchronous
generators. The implication is that WPPs do not provide primary frequency response, so the grid must
rely on conventional generators for it. Recently, wind turbine manufacturers have started offering
advanced control systems that provide synthetic inertial and governor-like response. This is discussed
in subsequent sections.
III.
WHAT IS GRID INTEGRATION OF WIND ENERGY?
Grid integration of wind energy is simply a collection of all activities related to connecting WPPs to the
grid. Activities are split into three categories based on when the activities occur. The first stage,
planning, includes activities that occur before a WPP is integrated. Physical connection encompasses
activities that occur during the physical connection of the wind farm to the grid. The final stage is
system operations, which are the activities that occur after the WPP is connected to the grid (Figure 2).
Figure 2: High-Level Activities of Grid Integration
Planning
•Power flow, short-circuit,
system stability studies
•System operations study
•Wind power interconnection
code
WPP= wind power plant.
Source: Asian Development Bank.
Physical
connection
•Build transmission from WPP
to substation
•Connection at substation
System operations
•Unit commitment
•Economic dispatch
•Wind energy forecasting
Grid Integration of Wind Power: Best Practices for Emerging Wind Markets
A.
7
Planning
There are two distinct types of planning activities related to grid integration: network-wide and
project-specific. Network-wide activities set the stage for all future WPPs; these include grid code
development, network-wide system integration studies with scenarios for different levels of wind
penetration, and system operations studies. Project-specific planning activities include system impact
studies done for a specific wind project. The methodologies for both types of studies are the same,
except for input data related to wind power installations. In a network-wide study, the input is wind
energy targets by region; while for a project-specific study, the input is WPP of the project.
1.
System impact studies
System impact studies for integrating wind energy are similar to studies performed when integrating
thermal power plants. Three types of impact studies are performed to determine the nature and extent
of impact: power flow, short-circuit, and power system stability (see Appendix I for details). In a
network-wide system, impact studies scenarios are created for various levels of variable renewable
energy (VRE) penetration over time; for example, 5% VRE in 5 years, 10% VRE in 8 years, and 15% VRE
in 12 years. In addition to determining the required upgrades, the cost of integration is also computed.
(It is the authors’ opinion that in most grids, the impact of 5% to 10% penetration of wind energy is
minimal.) Nevertheless it is a mistake to skip the impact study and assume that there is no impact
because a general rule cannot be prescribed, as the impact will depend on the specific grid under study.
An illustration of cost impact in various regional US grids is shown in Figure 4. A study in the United
Kingdom (International Energy Agency. 2005) concluded that with 5.3% penetration of wind, the lower
and upper estimate of cost are 1.3 to 2.1 €/MWh.
2.
Power system operations assessment
Power system operations are a collection of activities performed in multiple time frames (daily, hourly,
minutes-to-seconds) related to running the grid in a reliable and cost-effective manner. The total
demand in the power system must be balanced with the total generation from all sources at every
instant. This balance must happen for both active and reactive power. Traditionally, wind power was
considered nondispatchable, meaning when wind energy is produced, all of it must be absorbed by the
grid and the conventional generators must adjust around the timing and quantity of wind generation.
In developed wind energy markets, this traditional notion has changed, WPPs are now participants in
the dispatch process. Note that wind has “limited” dispatchability—WPP output can only be reduced
(not increased) from its current level of production (which is determined by current wind speed). The
benefits are reliable and cost-effective grid operations. To illustrate this, consider an extreme situation
in which the system operator is in a bind because of one, an imbalance in the grid—wind generation is
high, demand is low, and thermal generation is at its minimum; and/or two, transmission is
constrained—wind generation is high and transmission is reaching its thermal capacity. In such
situations, it may be cost-effective, from a system-wide perspective, to curtail wind energy production
(using for example active power control on turbines) until the constraints are resolved. The cost of
curtailing wind energy may be smaller than the cost of shutting down and restarting a thermal
generator in a short time period.
8
ADB Sustainable Development Working Paper Series No. 43
In emerging wind markets, it is imperative to perform an analysis to determine the impact of integrating
wind into the system operation’s unit commitment process and economic dispatch process. In a
flexible grid with small penetration of wind, wind-integrated dispatch may not be required. However in
an inflexible grid (see section V), wind-integrated dispatch may be required even for wind capacity as
low as 5% of peak load. Wind-integrated dispatch would require upgrade to the power system
operations process.
After assessment of systems operations, the primary upgrades are
(i)
incorporation of wind energy forecast to the unit commitment process and economic
dispatch process, where forecasts are required to predict the hourly variability of wind
energy production;
(ii) reduction of forecast lead time and dispatch interval; and
(iii) installation of active power control on wind turbines in order to limit production to a
set level.
If wind energy curtailment is substantial, then the grid can be increased flexibly using options like
shifting demand to off-peak hours, adding storage, and replacing the must-run inflexible generators
with flexible generators. This is discussed in more detail in Section V.
B.
Physical Connection
Physically, a WPP is connected to an existing high voltage transmission line in a substation. In order to
avoid the expense of building a substation, most WPPs choose to connect to an existing substation. In
the substation, the WPP line is connected to the medium voltage busbar, which is connected to a
transformer to step up the voltage to the high voltage transmission line. The medium voltage line from
the WPP to the substation is built by the WPP. The detailed design of the interconnection depends on
the conditions on the ground.
Within a WPP, a collection system is designed to connect the individual WTGs to the WPP substation.
Each WTG has a transformer that steps up voltage from 0.69 kilovolts (kV) (the typical output voltage
of a WTG) to medium voltage level (usually 11 kV or 33 kV) of the collection system. The medium
voltage collection system is normally buried underground. The substation of the WPP aggregates feeds
from the collection system and connects to the high voltage transmission line.
After the WPP is constructed and ready for commissioning, the operations engineers from the
transmission operator are involved in the commissioning phase to ensure that the WPP is safe to
operate and it produces energy of acceptable quality. After issues related to commissioning are
resolved and supervisory control and data acquisition connectivity is established, a request for initial
synchronization is made to system operations. The synchronization and subsequent testing process
depends on the transmission company’s processes and checklists. Primary tests performed include
(Jain 2010) (i) proper functioning of start-up, shutdown, and emergency shutdown; (ii) proper
functioning of switch gear in response to a variety of fault conditions; (iii) harmonics, flicker, and other
quality parameters are within limit; and (iv) successful communication of data from WPP to
supervisory control and data acquisition system.
Grid Integration of Wind Power: Best Practices for Emerging Wind Markets
C.
9
System Operations
After the WPP is physically connected to the transmission line, the attention of grid integration turns to
system operations, which integrate the production of WPPs into the grid operations. The goal of grid
operations is to commit and dispatch all generation units in the system with the objective of meeting
demand in a reliable and economical manner. Wind energy forecasting is an important enabler in order
to meet these objectives, given the variability and uncertainty of wind resources. Day-ahead hourly
wind energy forecasts are used in the unit commitment process and intraday updates to forecasts are
used in the economic dispatch process. The unit commitment process and economic dispatch process
are described in the next section.
IV.
WHAT IS THE IMPACT OF A WIND POWER PLANT ON A GRID?
Integrating any new power plant has an impact on the existing grid and the incumbent generators.
In the following, impact on multiple time scales will be described i.e., day-ahead, hour-to-hour,
minute-to-minute, and seconds-to-sub-seconds.
A.
Day-Ahead Unit Commitment Process
System operators create a daily unit commitment process plan based on forecasts of load and
characteristics (maximum and minimum production, ramp rate and cost) of generation plants. The
output of a unit commitment process is a day-ahead hourly “on/off” scheduling of generation units. It
however does not commit to the level of production. The unit commitment process is therefore a
“cost-effective combination of generating units to meet forecasted load and reserve requirements,
while adhering to generator and transmission constraints.” When a WPP is integrated to a grid, it is
usually required to provide a day-ahead hourly forecast of energy production. Basically, the unit
commitment process incorporates the hourly wind energy forecast by adjusting the hourly on/off
status of the most expensive generators. If there is sufficient wind generation for one or more hours,
then it may displace the most expensive generation on the grid.
Uncertainty of wind and demand forecasts is an issue that must be managed by a unit commitment
process. It is best explained with an example. If the actual output of WPP at a specific hour is 30% less
than predicted in the day-ahead forecast, then during the hour, the committed generators will have to
make up the difference. If the maximum production levels of the committed generators cannot make
up the 30% loss of wind, then spinning reserves, quick start units, or short-term purchases will
compensate for the energy shortfall.
B.
Economic Dispatch Process
System operators set the level of production of the committed generators an hour ahead of delivery.
The level of production of generators is optimized to minimize the total cost of generation while
meeting all the constraints of reserve requirement, generator properties, and others. This is called
economic dispatch. In most grids, wind energy has the highest priority (lowest marginal cost), therefore
the process dispatches all the wind energy that is forecast for the hour while optimizing the economics
of the other generators. Hence when wind generators are producing energy, the level of production
allotted to the most expensive generators is less.
10
ADB Sustainable Development Working Paper Series No. 43
C.
Illustration of Variability and Uncertainty of Wind Energy
To illustrate how variability 2 of wind energy impacts dispatch, consider diurnal profile of wind in which
wind is weak from 7 a.m. to 10 p.m., and strong from 10 p.m. to 7 a.m. Consider a demand profile in
which demand is strong from 7 a.m. to 10 p.m. In this situation, as demand picks up in the morning,
wind output is decreasing; which implies that other generators in the grid must ramp up faster. Similarly
in the evenings, load is decreasing while wind output is picking up, which implies that the other
generators in the grid must ramp down faster. So when variability of demand and wind are not
synchronized, then the other generators on the grid must have the ability to ramp up or down faster. As
wind penetration increases, the ramp rates may become too high for existing generators to handle, in
which case additional spinning reserves may be required. With lower penetration of wind, there may be
no change to the unit commitment process, while the economic dispatch process manages the ramp
rates. As penetration increases, the unit commitment process would have to commit more generators
to support high ramp rates.
To illustrate how uncertainty of wind impacts dispatch, consider a grid with total load of 1,000 MW and
two scenarios—5% and 20% wind power penetration. In the first scenario with 5% penetration,
consider a case in which the WPP output is forecasted to be 50 MW (maximum production) and
actual WPP output is 30% below forecast. In this case, the committed generators must make up for
15 MW (30% of 50 MW) in shortfall of power, which is 1.58% (=15 MW/950 MW) of the committed
capacity. Most grids should be able to manage this uncertainty using the cushion available in the
dispatch plan—the difference between maximum and dispatched capacity of the committed
generators. The reason is most grids are designed to manage 3% to 5% uncertainty in load.
Consider the second scenario in which the total load is 1,000 MW; WPP output is forecast to be
200 MW (maximum production); and actual WPP output is 30% below forecast. In this scenario the
committed generators must make up for 60 MW (30% of 200MW) in shortfall of power, which is 7.5%
(=60 MW/800 MW) of the committed capacity. Most grids may not be able to manage this level of
uncertainty with the existing generation resources. In such scenarios, possible solutions are additional
quick start units, or exchange of energy with neighboring grids.
As the level of wind energy penetration goes up, the reserve requirement goes up and the value of
reducing uncertainty of wind energy forecasts in the unit commitment process and the economic
dispatch process goes up. Therefore, management of wind energy uncertainty in the unit commitment
process and the economic dispatch process has a prominent role. Moving from day-ahead hourly wind
energy forecasts to intraday (for example, update to forecasts every 2 hours) wind energy forecasts
would reduce uncertainty.
D.
Implications of Wind Energy on Conventional Generators and Transmission
Besides the impact on the unit commitment process and the economic dispatch process, wind energy
has other impacts on the grid.
2
The energy output is variable because the wind speed is variable. The variability is on all time scales—second-to-second,
minute-to-minute, hour-to-hour, day-to-day, month-to-month, and year-to-year variability.
Grid Integration of Wind Power: Best Practices for Emerging Wind Markets
1.
11
There are three primary implications on existing thermal and hydro generation units in
the grid.
(i)
Higher ramp rates. In the worst case scenario, the thermal generation units will have to
support the sum of maximum demand ramp rate (d MW/minute) and maximum wind
generation ramp rate (w MW/minute), which is (d+w) MW/minute. If the sum of ramp
rates is high and cannot be supported by the existing generation units on the grid, then
restrictions may be placed on WPP to ramp up or ramp down production of individual
turbines in a staggered manner.
(ii) Deeper cycling. As described earlier, output of WPPs may be treated as negative load in
the grid, which results in deeper valleys in the net load profile, which then results in
deeper cycling of existing generation units.
(iii) More frequent start and stop. During the high wind season, supply-demand balance at
the unit commitment process level may force displacement of the most expensive
thermal power plant in favor of WPP, which will result in higher frequency of start and
stop of thermal generators.
2. At high wind energy penetration levels, the inertia of the grid system decreases (Seyedi and
Bollen 2013, Miller et al. 2010) because the percentage of synchronous generation is
reduced. As a result, the ability of the grid to respond to loss of large production unit
declines. Traditionally conventional synchronous generators (high inertia) provide
frequency regulation. If frequency regulation is an issue during high wind penetration, then
WPP may be required to provide synthetic inertia, which is accomplished by a control
system that transfers mechanical inertia from a rotor to the electrical side in Type 3
double-fed induction generator and Type 4 full-power converter-based turbines. If the
WTGs are unable to provide inertial response, then synchronous generators would have to
stay online to maintain reliability, which may lead to curtailment of wind energy.
3. To manage small-signal stability, which is response to variability and uncertainty of wind in
the second-to-second time frame, inertial response may be required, as described above.
4. To manage in the tenths of seconds to minute time frame, upgrades to governors and
controls of current thermal and hydro units may be required.
5. To manage in the minute time frame, upgrade to automatic generation control may be
required in addition to more spinning reserves.
6. The flow of electricity in the grid is impacted by introduction of wind energy. More
electricity will flow through some nodes, therefore transmission line, substations, and
protection systems may need to be upgraded.
V.
BEST PRACTICES FOR WIND INTEGRATION IN EMERGING MARKETS
Wind integration has not been easy, especially in grids where it was an afterthought. In the East Inner
Mongolia region of the People’s Republic of China, 23% of wind power was curtailed in 2011 due to
unavailability of transmission to transport wind energy to load centers. In the US, the Electric Reliability
Council of Texas system faced 17.1% curtailment in 2009, which was brought down to 2.5% in 2012.
12
ADB Sustainable Development Working Paper Series No. 43
These two examples are extreme cases, and significant experience has been gained from these and
other cases. The insights are generalized as best practices below, with a focus on emerging wind energy
markets.
A.
Transmission from Resource-Rich Areas to Load Center
If transmission is a bottleneck, then policies must address the “chicken-or-egg” dilemma, which is,
Should transmission be built in anticipation of WPP or should WPPs be built in anticipation of
transmission? The policy that has been successful in Texas is to define wind corridors or preferred
zones that are earmarked for investments in transmission and other infrastructure. Wind energy
corridors are regions screened for wind resources, land availability, infrastructure, and other factors
with the goal of focusing the efforts of all agencies to the identified regions. Such corridors can channel
limited resources to one region or a few regions, thereby leveraging transmission and other
infrastructure investment in a corridor to support multiple WPPs. This would be in contrast to
suboptimal, project-specific approach, which would require the utility to plan and invest in upgrades
per project in the region of the project.
In emerging wind energy markets, transmission may not be a bottleneck for the first few projects
because the existing lines have sufficient unutilized capacity. However this must be checked through a
power flow study, which will go beyond checking adequacy of transmission line capacity to determining
requirements for reactive power compensation on transmission lines greater than 100 kilometers
(Matevosyan 2005) to ensure voltage and transient stability. In most cases, there is a reversal of
current flow in an existing line, which requires reactive power compensation to keep the voltage at POI
with limits permitted by the grid code.
B.
Responsive System Operations
1.
Reducing forecast lead time and dispatch interval
In most emerging wind energy markets, day-ahead hourly schedules (forecast lead time is 1 day and
dispatch interval is 1 hour) are created based on load and wind energy forecasts. The daily schedule of
generators is firmed up on the previous day, in the late afternoon or early evening. In the schedule,
output is set for each hour for each generator based on hourly net-load. Since wind and solar are
variable “nondispatchable” generators with near-zero marginal cost, the wind and solar energy
forecasts are used as schedules, and all other generators are scheduled around them.
Real time imbalance between net-load and generation is cleared by redispatch and ancillary services
such as regulating services. Regulating reserve is the most expensive option. Redispatch should be
utilized as much as possible. However, some electric utilities perform dispatch once an hour, which
makes grids inflexible (Holttinen 2013). If the dispatch interval is reduced to 15 minutes, then the
dispatched generation, with the capability to adjust generation output, can respond to changes every
15 minutes, leaving regulating reserves to manage much smaller intra-15-minute variability and
uncertainty. This unlocks flexibility available in dispatched generators, and enables a larger amount of
variability and uncertainty to be managed in an economical manner.
One such example is from the western United States with 37 balancing areas and 160 GW of
generation, and scenario of 23% variable generation (Holttinen 2013). In this example, when the
balancing area is increased and dispatch intervals are reduced from 60–40 to 10–10, the average total
regulation required drops from 9 GW to 1 GW. Here 60–40 means 60 minutes dispatch interval and
Grid Integration of Wind Power: Best Practices for Emerging Wind Markets
13
40 minutes forecast lead time; while 10–10 means 10 minutes dispatch interval and 10 minutes
forecast lead time.
Sub-hourly forecast lead time and dispatch intervals can significantly improve flexibility (Cochran et al.
2012). In most cases, this is the cheapest form of flexibility that can be exploited. Therefore it is the
most relevant for emerging wind markets. The exact forecast lead time and dispatch interval do not
have to be set on day one. The system operator may choose to start with larger intervals and
progressively move to shorter intervals. The intervals depend on likely amount of curtailment,
forecasting accuracy of wind energy production, and flexibility of other generators. In the US, several
independent system operators with high penetration of wind have adopted 15-minute scheduling and
dispatch intervals.
2.
Active power control
As wind penetration increases, reducing forecast lead time and dispatch interval may not be sufficient
to balance supply and demand in all situations. The most challenging situations are when wind energy
production is higher than the difference between the load and sum of minimum level of production of
dispatched generators. Seasons play an important role in a unit commitment process and the
economic dispatch process. For example, in winter, heat requirements may warrant combined heat and
power plants to run at much higher capacity levels, leaving limited room for wind energy. In the rainy
season, hydro power plants may be forced to run at higher capacity, leaving limited room for wind
energy. In such situations active power control features of wind turbines can be useful. System
operators may dispatch WPP to a set level of power production, which is below the forecasted level.
That is, wind is no longer a nondispatchable resource, but is a participant in economic dispatch with
limitations (wind production cannot be increased beyond the available wind resource). Note that in
this situation, wind energy is being curtailed.
There are three key considerations (Ela et al. 2014) when implementing wind-integrated dispatch:
(i) WPP must be adequately compensated for reduced production and additional investment in active
power control; (ii) the new controls of WPP should not adversely impact the stability of the grid; and
(iii) the new controls should not increase the loading of turbines, so as not to adversely impact the life
of the turbines.
3.
Flexible generation
Inertial response and governor response (described in the previous section), as well as reserves,
provide flexibility to the grid in responding to supply–demand imbalances. Reserve generators with
high ramp rates and low minimum operating capacity are flexible generators because these can be
dispatched within minutes and can change production levels rapidly. Traditionally, quick response was
provided by gas peaking plants, hydropower, pumped storage, and interconnection to neighboring
grids. More recently, options like refurbishing of conventional power plants for flexible operations,
utility-size battery storage, compressed air storage, flywheels, and others are used to enhance
flexibility. Storage is a perfect flexible resource in terms of technical capabilities, but is one of the most
expensive forms of flexibility. It can store and release energy (over longer periods of time), power (rapid
exchange over seconds and minutes), or both. Pumped hydro storage and compressed air energy
storage are examples of energy storage used for storing energy that would otherwise be curtailed.
Flywheel is used for storing power, which is used for frequency regulation and ramping over short
intervals. Batteries are used for both energy and power storage.
14
ADB Sustainable Development Working Paper Series No. 43
4.
Flexible demand
Load or demand can provide flexibility to the grid in a variety of ways. Instead of generators responding
to imbalance between supply and demand, in this case the demand responds. In emerging wind
markets, very little demand flexibility is utilized. Demand flexibility can take two forms: load balancing
and load shifting. Demand response (DR) is a form of load balancing, in which load is “dispatchable.”
Dispatchable loads are customers who possess the flexibility to postpone their operations by a few
hours, or have the flexibility to reduce power consumption for a few hours. Examples are wastewater
treatment plants, ice making plants, chillers, and other heating ventilation and air conditioning
equipment, and others. Although this is not practiced much in the developing wind markets, it is
starting to play a bigger role in grids with significant wind power penetration.
Load shifting is the other flexibility in demand that deals with the changing behavior of large consumers
of electricity. This is achieved by time-of-day tariff, which provides an incentive to large consumers of
electricity to shift load from times of high tariff (peak demand) to times of low tariff (off-peak demand
and peak wind energy production).
5.
Grid code for wind energy integration
Grid code for interconnection is a rulebook that specifies properties that generators and other
equipment should satisfy in order to connect to the grid. The goal of the grid code is to ensure reliable
and safe operation of the grid.
Variable generators like wind turbines and solar photovoltaic cells did not exist when the original grid
codes were created, therefore a grid code refers only to synchronous generators like diesel, steam,
hydro, or combined heat and power. Often in emerging renewable energy markets, the grid code is
updated to explicitly account for wind and solar generation only after a few installations. In an
emerging wind energy market, the existing grid code must be evaluated for compatibility with variable
generation and other characteristics of WTGs like asynchronous generators and full power conversion.
Appendix II contains an outline of a wind power-specific grid code.
From a planning perspective, the grid code for interconnection of wind power plant should be a top
priority. Without a grid code specific to WPP, a wind developer is not obligated to provide “high quality”
power output or appropriate fault ride through response.
Sometimes, in the zeal to create a rigorous grid code, the regulators and grid operators create a grid
code that is too stringent and does not reflect the realities of the grid. For instance, if the grid code
specifies voltage and frequency bands that are too narrow, then the WPP may be forced to trip very
often; or if the grid code has conditions that are too onerous, then the cost of WPP may be too high for
a feasible project. A balance must therefore be reached between specifications that are too loose and
too stringent. It should also be noted that grid code is not a static document; it must be updated
routinely in response to changes in the grid and new technologies.
6.
Comprehensive planning process
The next best practice relates to planning—long-term and project-specific short term. The previous
five best practices highlight the enormity of the task and the need for long-term sustained effort.
Institutionalizing a tiered planning process, which starts with an integrated energy master plan that
includes wind energy and ends with wind project-specific system impact studies, can lead to faster
Grid Integration of Wind Power: Best Practices for Emerging Wind Markets
15
growth of the wind industry. Often the initial wind energy projects are considered novel and barely
register a blip in planning processes. In emerging wind energy markets, grid integration occurs on an ad
hoc basis. There is a general feeling that interconnection of the first few wind projects will have an
insignificant impact on the grid. As a consequence, a comprehensive impact analysis, which includes
system operations, is not performed.
It is however imperative that the interconnecting utility integrate wind energy annual targets into the
long-term energy master plan. After the wind energy targets are modeled in network-wide power flow
and dynamic stability models, then impact of wind energy on the network can be analyzed.
Subsequently, longer-term plans can be developed for enhancing the flexibility of the grid in
anticipation of larger penetration of variable renewable energy.
At the other end is wind project-specific planning. As part of wind project licensing, processes must be
developed that require the interconnecting utility to perform a system impact study for every wind
energy interconnection request. This will force the utility to assess the impact of a WPP on the grid
based on power flow, short-circuit, dynamic stability studies, and system operations studies; and
consequently make necessary changes to the grid prior to integration.
Figure 3: Tiered Planning for Grid Integration
Source: Asian Development Bank.
VI.
HOW MUCH DOES GRID INTEGRATION COST?
Integration costs may be divided into three categories: (i) transmission extension and related
reinforcement, (ii) balancing of increased volatility in the grid, and (iii) maintaining the adequacy
(ability to cover peak load). For easier comparison, the first cost category is not included. Figure 4
summarizes the results of studies on wind integration cost (excluding cost category one) for different
levels of wind power capacity in the grid for different grids. The studies in the figure present the cost of
16
ADB Sustainable Development Working Paper Series No. 43
integration in six markets. 3 Four of the six markets have wind integration costs that are below $5 per
MWh for wind capacity penetration as high as 30%. Across all the studies, the cost of wind integration
was 1% to 14% of the total levelized busbar cost of wind electricity in 2010.
Recent studies in the US show that the cost of grid integration is even lower. Studies prior to 2008 had
estimated integration costs up to $5/MWh. By 2013, Electric Reliability Council of Texas, which has the
highest penetration of wind, was reporting integration cost of $0.5/MWh, primarily due to operating
reserve requirements.
In emerging wind markets with low penetration of wind energy, existing transmission lines are used.
However, as congestion on transmission lines increase, new transmission lines may be required. New
transmission line cost is not allocated to a wind project, because the lines are used by multiple projects
and multiple types of power plants. Instead, usage-based costs (per MWh of energy transported)
are paid to transmission line operators, and these costs are reflected in operation and maintenance
cost of a WPP.
Figure 4: Studies on Wind Integration Costs versus Wind Capacity Penetration
in Various Regions of the United States
Capacity Penetration (%)
Source: Rocky Mountain Institute.
Other estimates in the literature suggest that at a 20% share of average electricity demand, wind
energy balancing costs range from $1/MWh to $7/MWh (IEA 2011).
The Grid Integration of Variable Renewables project of the International Energy Agency has simulated
the cost of a hypothetical case that has 45% variable renewable energy in the grid (IEA 2014a or b?).
Two scenarios are analyzed: first, 45% variable renewable energy is added overnight (base-case);
3
Please see citations contained in the summary of Rocky Mountain Institute for detailed assumptions associated with the
cost estimate.
Grid Integration of Wind Power: Best Practices for Emerging Wind Markets
17
second, it is added over time in an orderly and coordinated transformation. In the first scenario, the
total system costs increase by as much as $33/MWh; the total levelized cost rises from $86/MWh to
$119/MWh. In the coordinated transformation scenario, a larger number of flexible power plants are
deployed and the grid infrastructure is shared and managed across all of the generation.
The economics of adding flexibility to the grid are system-specific. Experiences across countries and
across independent system operators within a country to enhance flexibility have revealed a picture
of relative economics as described in Figure 5. The most inexpensive form of flexibility is obtained
from system operations—grid codes that require WPP to be grid friendly; wind energy forecasting;
sub-hourly scheduling and dispatch; sharing of reserves and enhanced coordination with neighboring
balancing areas; and as a last resort, strategic curtailment. In competitive electricity markets,
market-based solutions like nodal pricing and ancillary services pricing can increase grid flexibility
because the price signals invite more investment to the ancillary service market. Enhancing load
flexibility through demand response programs is the other more economical option; however, load
flexibility through involuntary load shedding is expensive. Currently, flexibility is provided by
conventional generators, primarily by hydropower and gas power plants; coals power plants may be
refurbished to provide a higher amount of flexibility. Additional transmission and variety of network
management devices and strategies can enhance flexibility by removing bottlenecks associated with
power flow. Relatively, the most expensive form of flexibility is from battery-based storage, while
thermal and pumped-hydro storage are more economical.
Figure 5: Relative Cost of Various Options to Increase Flexibility
of the Grid with Varying Wind Percentages
CCGT = combined cycle gas turbine, CT = combustion turbine, RE = renewable energy.
Note: Six classes of interventions are examined, as labeled on the X axis.
Source: Cochran et al. 2014.
18
ADB Sustainable Development Working Paper Series No. 43
In Figure 5, as evidenced from the above description, physical flexibility (the three interventions on the
right of the chart) is expensive, while operational practices and market design are relatively economical.
VII.
CONCLUSIONS
Experience around the world has demonstrated that wind energy can be integrated into power systems
reliably and economically.
In emerging wind energy, grid integration is an afterthought. This has resulted in a variety of issues: poor
power quality, large voltage variations at POI, delayed interconnection after the wind plant is
commissioned, and significant curtailment. The result is investments in wind energy projects slow
down.
Grid integration has three components: technical, planning, and policy. Just because a conventional
generator can operate in a more flexible manner does not mean it will operate flexibly; policies must be
in place to incentivize the various parts of the grid to release locked flexibility.
Most emerging wind energy countries have a single buyer market, in which the buyer runs large
fiscal deficits. In this context, if the wind power plants are perceived as adding to the deficit of the
buyer, then the buyer has no incentive to solve the grid integration problems, irrespective of the
environmental benefits. Therefore, incentivizing the buyer and other enabling grid integration policies
plays an important role.
The following best practices for emerging wind energy markets should lead to effective and
cost-efficient grid integration of wind projects.
(i)
Transmission from resource rich areas to load centers. Since transmission projects have a
longer lead time than wind projects, new transmission or upgrade to transmission must be
planned ahead of time. A policy of defining wind energy corridors and focusing
transmission investment in a corridor is preferable to a project-specific approach. Note
that this approach requires the government to have a high level of confidence in both the
wind resource and near-term uptake in wind project development. This approach has
been adopted in Mannar, Sri Lanka, and through the CREZ initiative in Texas. 4
(ii) Responsive systems operations. Upgrading the scheduling and dispatching process for all
generators to sub-hour dispatching intervals, incorporating sub-hour forecast lead time
for wind energy forecasts, and expanding the balancing areas are process-related
improvements that offer the cheapest form of grid flexibility.
(iii) Flexible generation. Variability in a grid in not new; it is managed by flexible generation.
Wind energy introduces additional variability that must be served by generators with
higher ramp rates and deeper cycles. If the dispatched generators are unable to provide
the ramp rate, then additional generators should be dispatched at partial load. To support
high wind penetration, wind turbines should have capability to provide inertia and
governor-like response.
4
H. Smith. 2014. Texas CREZ Project Completed and Already Responsible for Expanded Wind Energy Development.
10 February. http://www.texasvox.org/texas-crez-project-wind-energy/
Grid Integration of Wind Power: Best Practices for Emerging Wind Markets
19
(iv) Flexible demand. Wind curtailments occur during periods when wind production is higher
than the difference between load and the sum of minimum operating levels of all
dispatched generators. Load shifting to such periods from periods of peak load can
enhance the ability of the grid to absorb higher amount of variable energy.
(v) Comprehensive planning process. There is no one-size-fits-all solution that applies to all
grids. A network-wide grid integration impact assessment that examines the impact on
the grid of various percentages of wind energy penetration as a function of time is
required to determine the grid integration issues and solutions.
(vi) Grid code for wind integration. A grid code for connecting wind plants is essential because it
specifies the minimum technical criteria that all wind farms seeking connection to the grid
shall satisfy at POI, and the operating conditions that all wind farms shall comply with for
safe and reliable operations of the grid.
Wind integration can be further facilitated and integration costs further reduced by efficient grid
operating procedures such as large or coordinated balancing areas, fast-interval generation scheduling
and dispatch, setting wind generator schedules as close as possible to the dispatch time to minimize
forecast errors, and use of wind power forecasting.
REFERENCES
American Wind Energy Association. Wind Energy, Backup
http://www.awea.org/Issues/Content.aspx?ItemNumber=5454
Power,
and
Emissions.
J. Cochran, L. Bird, J. Heeter, and D. J. Arent. 2012. Integrating Variable Renewable Energy in Electric
Power Markets. NREL/TP-6A00-53732. Boulder: National Renewable Energy Laboratory.
J. Cochran, M. Miller, O. Zinaman, M. Milligan, D. Arent, and B. Palimintier. 2014. Flexibility in the
21st Century Power Systems. NREL/TP-6A20-61721. Boulder: National Renewable Energy
Laboratory.
E. Ela, V. Gevorgian, P. Fleming, Y. Zhang, M. Singh, E. Muljadi, A. Scholbrook, J. Aho, A. Buckspan,
L. Pao, V. Singhvi, A. Tuohy, P. Pourbeik, D. Brooks, and N. Bhatt. 2014. Active Power Controls
from Wind Power: Bridging the Gaps. NREL/TP-5D00-60574. Boulder: National Renewable
Energy Laboratory.
H. Holttinen, A. Touhy, M. Milligan, E. Lannoye, V. Silva, S. Muller, and L. Soder. 2013. The Flexibility
Workout. IEEE Power and Energy Magazine. pp. 53–62.
International Energy Agency. 2005. Variability of Wind Power and Other Renewables: Management
Options and Strategies. Paris.
———. 2014a. Harnessing Variable Renewables: A Guide to the Balancing Challenge.
https://www.iea.org/Textbase/npsum/Harness_Renewables2011SUM.pdf
———. 2014b. The Power of Transformation: Wind, Sun and the Economics of Flexible Power Systems.
http://www.iea.org/Textbase/npsum/GIVAR2014sum.pdf
P. Jain. 2010. Wind Energy Engineering. New York: McGraw-Hill.
G. Jordan and S. Venkataraman. 2012. Analysis of Cycling Costs in Western Wind and Solar Integration
Study. Boulder: National Renewable Energy Laboratory.
D. Lew, G. Brinkman, N. Kumar, S. Lefton, G. Jordan, and S. Venkataraman. 2013. Finding Flexibility:
Cycling the Conventional Fleet. IEEE Power and Energy Magazine. 11(6, December): pp. 20–32.
J. Matevosyan. 2005. Wind Power in Areas with Limited Transmission Capacity. In Wind Power in Power
Systems. West Sussex: John Wiley & Sons. pp. 433–459.
A. McCrone, E. Usher, V. Sonntag-O’Brien, U. Moslener, and C. Grüning. 2014. Global Trends in
Renewable Energy Investment 2014. Frankfurt School–UNEP Collaborating Centre, United
Nations Environment Programme (UNEP), and Bloomberg New Energy Finance.
N. Miller. 2013. GE Wind Plant Advanced Controls. In First International Workshop on Grid Simulator
Testing of Wind Turbine Drivetrains: Workshop Proceedings. Boulder: National Renewable Energy
Laboratory.
N. Miller, K. Clark, and M. Shao. 2010. Impact of Frequenct Responsive Wind Plant Controls on Grid
Performance. In Proceedings of the 9th International Workshop on Large-Scale Integration of Wind
References
21
Power into Power Systems as Well as on Transmission Networks for Offshore Wind Power Plants.
Quebec City. 18–19 October 2010, Québec City, Canada.
N. Miller, C. Loutan, M. Shao, and K. Clark. 2013. Emergency Response. IEEE Power and Energy
Magazine. 11(6, November/December):63–71.
National Renewable Energy Laboratory. Western Wind and
http://www.nrel.gov/electricity/transmission/western_wind.html
Solar
Integration
Study.
National Renewable Energy Laboratory. 2013. NREL Calculates Emissions and Costs of Power
Plant Cycling Necessary for Increased Wind and Solar in the West. 24 September.
http://www.nrel.gov/news/press/2013/3299.html
Rocky Mountain Institute. Total Wind Integration Costs for Different Capacity Penetrations.
http://www.rmi.org/RFGraph-Total_wind_integration_costs_capacity_penetrations
M. Seyedi and M. Bollen. 2013. The Utilization of Synthetic Inertia from Wind Farms and Its Impact on
Existing Speed Governors and System Performance. Elforsk. Rapport 13:02.
J. Slootweg and W. King. 2005. Impacts of Wind Power on Power System Dynamics. In Wind Power in
Power Systems. New York: John Wiley & Sons. pp. 629–640.
US Department of Energy. 2015. Wind Vision: A New Era for Wind Power in the United States.
Washington, DC.
Grid Integration of Wind Power
Best Practices for Emerging Wind Markets
Issues with grid integration of wind energy has led to curtailment of wind power, delay in interconnection for
commissioned wind projects and/or denial of generation permit. This report describes the impact of wind power
on the grid, methods to analyze the impact and approaches to mitigate the impact. Countries like Denmark,
Germany, Spain, and regions within the United States like Texas and Colorado have achieved high penetration
of wind energy with modest changes to the grid. The key to high penetration of wind energy has been flexible
grids. The report outlines the lessons learned with a focus on applicability to emerging wind energy markets.
GRID INTEGRATION
OF WIND POWER
About the Asian Development Bank
ADB’s vision is an Asia and Pacific region free of poverty. Its mission is to help its developing member countries
reduce poverty and improve the quality of life of their people. Despite the region’s many successes, it remains
home to the majority of the world’s poor. ADB is committed to reducing poverty through inclusive economic
growth, environmentally sustainable growth, and regional integration.
BEST PRACTICES FOR EMERGING
WIND MARKETS
Based in Manila, ADB is owned by 67 members, including 48 from the region. Its main instruments for helping
its developing member countries are policy dialogue, loans, equity investments, guarantees, grants, and
technical assistance.
Pramod Jain and Priyantha Wijayatunga
NO. 43
April 2016
ASIAN DEVELOPMENT BANK
6 ADB Avenue, Mandaluyong City
1550 Metro Manila, Philippines
www.adb.org
ADB SUSTAINABLE DEVELOPMENT
WORKING PAPER SERIES
ASIAN DEVELOPMENT BANK
Download